Forensic genetics and the prediction of race: what is the problem?

A new wave of innovations in forensics seeks to support criminal investigations by making inferences about the racial or ethnic appearance of as yet unknown suspects using genetic markers of phenotype or ancestry. This paper argues that to grasp fully the potentials of these innovations they must be understood both in the context of established patterns of police-minority relations and as part of significant changes in the use of ‘race’ as an object of knowledge in science, policy, and politics. Socio-technical developments offer new means of identification through geneticization, datafication, and visualization and heighten the visibility and valorisation of racial difference. Elements of this are already evident in existing national police forensic DNA databases whose operation, outcomes, and accompanying ethical frames are racialized in varied ways. By openly mobilising race and ethnicity, however, predictive techniques raise new questions about the validity, interpretation, dissemination, and application of results. Examination of existing use by the police and public of suspect descriptions shows the enduring power of common sense visual and linguistic understandings of race and appearance. That very power makes it hard to transition effectively from moments of collective stigmatization to the identification of individual suspects.